When you query chunks of data using "GET", you will get by default five entries in a chunk -- see http://cumulocity.com/guides/reference/rest-implementation#-a-name-paging-a-query-result-paging. This is done to not overload smaller or embedded clients.
If you have a more powerful client, such as a microservice, we recommend increasing the page size as much as possible. This will, instead of running many requests with small results, run a few larger requests with larger results, with a better overall performance. For example, in the Java client, use the "get"-Method of PagedCollectionResource that permits you to pass a pageSize parameter as first argument, and use, for example, the value 2000 to set the maximum page size.